def process(self):
""" parse the Golm Metabolome Database standards library.
such as
those found at
http://csbdb.mpimp-golm.mpg.de/csbdb/gmd/msri/gmd_msri.html
"""
parent = self.parent
parent.mm.setTableWeak('mpimpclass')
# make sure that the necessary tables are in the db
un = mask({}, parent.mm.idpatterns)
l = parent.getLine()
while l:
# check and send off gathered data
if re.match('^Name:', l):
if un.nid(parent.mm) > 1:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
x = re.findall('Name: (.+)', l)
mpimpid = x[0].strip()
un.append('mpimp', mpimpid, parent.confid,
parent.sourceid)
# get synonyms from the title, the other synonyms are
# just to full of crap
spam = mpimpid.split('_')
if spam[5].strip() != '':
un.append('synonym', spam[5], parent.confid,
parent.sourceid)
if re.match(".*\([0-9]+[tT][mM][sS]\)$",
spam[5]):
syn = re.findall("(.*)\([0-9]+[tT][mM][sS]\)$",
spam[5])[0].strip()
un.append('synonym', syn,
parent.confid, parent.sourceid)
# ***** gather info *****
# mpimp class
x = re.findall('Synonym: Metabolite \(class\):(.+)', l)
if x:
un.append('mpimpclass', x[0].strip(),
parent.mm.confidence['weak'],
parent.sourceid)
# cas 1
x = re.findall('CASNO: (.+)', l)
if x:
un.append('cas', x[0].strip(), parent.confid,
parent.sourceid)
# cas 2
x = re.findall('CAS\|(\d{1,7}-\d{2}-\d{1})_\(.+\)', l)
if x:
un.append('cas', x[0].strip(), parent.confid,
parent.sourceid)
# kegg
x = re.findall('KEGG\|([A-Z][0-9]{5})_.*\(.+\)', l)
if x:
un.append('kegg', x[0].strip(), parent.confid,
parent.sourceid)
l = parent.getLine()
# send of the last mask if we have one
if un.nid(parent.mm) > 1:
parent.setMask(un)
def process(self):
parent = self.parent
ll = True
while ll:
ll = parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = parent.mm.getMask(tmp, weak=True)
if not tmp2:
continue
else:
un = tmp2[0]
if self.cs.fixInchi(un):
parent.setMask(un, setass=False)
tmp2 = parent.mm.getMask(tmp, weak=True)
if tmp2:
un = tmp2[0]
try:
csMasks = self.cs.getCsMasks(un, parent.mm)
except:
pdb.set_trace()
if parent.mm.debug:
print("#COMMENT mask " + str(ll) + " csid " + str(list(csMasks.keys())))
cmsk = mask({})
for csid in list(csMasks.keys()):
csMasks[csid].setAllAssoc(parent.mm.addAss())
cmsk.merge(csMasks[csid])
parent.setMask(cmsk, setass=False)
def process(self):
ll = True
self.parent.mm.setTableWeak('formula')
while ll:
ll = self.parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = self.parent.mm.getMask(tmp)
if not tmp2:
continue
else:
un = tmp2[0]
try:
chebiMasks = self.chebip.getChebiMasks(un, self.parent.mm)
except:
pdb.set_trace()
if self.parent.mm.debug:
print("#COMMENT mask " + str(ll) + " chebi " +
str(list(chebiMasks.keys())))
chMask = mask({})
for ch in list(chebiMasks.keys()):
chebiMasks[ch].setAllAssoc(self.parent.mm.addAss())
chMask.merge(chebiMasks[ch])
self.parent.setMask(chMask, setass=False)
def process(self):
parent = self.parent
weakTables = ['formula', 'weight', 'totalcharge', 'xlogp', 'hbonddonor',
'hbondacceptor', 'heavyatom', 'tpsa']
list([parent.mm.setTableWeak(x) for x in weakTables])
ll = True
while ll:
ll = parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = parent.mm.getMask(tmp, weak=True)
if not tmp2:
continue
else:
un = tmp2[0]
try:
pcMasks = self.pc.getPubchemMasks(un)
except:
print("#COMMENT unknown error")
continue
if parent.mm.debug:
print("#COMMENT mask " + str(ll))
msk = mask({})
for pcmsk in pcMasks:
pcmsk.setAllAssoc(parent.mm.addAss())
msk.merge(pcmsk)
parent.setMask(msk, setass=False)
def process(self):
ll = True
self.parent.mm.setTableWeak('formula')
while ll:
ll = self.parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = self.parent.mm.getMask(tmp)
if not tmp2:
continue
else:
un = tmp2[0]
try:
chebiMasks = self.chebip.getChebiMasks(un, self.parent.mm)
except:
pdb.set_trace()
if self.parent.mm.debug:
print("#COMMENT mask " + str(ll) + " chebi " + str(list(chebiMasks.keys())))
chMask = mask({})
for ch in list(chebiMasks.keys()):
chebiMasks[ch].setAllAssoc(self.parent.mm.addAss())
chMask.merge(chebiMasks[ch])
self.parent.setMask(chMask, setass=False)
def process(self):
parent = self.parent
ll = True
while ll:
ll = parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = parent.mm.getMask(tmp, weak=True)
if not tmp2:
continue
else:
un = tmp2[0]
if self.cs.fixInchi(un):
parent.setMask(un, setass=False)
tmp2 = parent.mm.getMask(tmp, weak=True)
if tmp2:
un = tmp2[0]
try:
csMasks = self.cs.getCsMasks(un, parent.mm)
except:
pdb.set_trace()
if parent.mm.debug:
print("#COMMENT mask " + str(ll) + " csid " +
str(list(csMasks.keys())))
cmsk = mask({})
for csid in list(csMasks.keys()):
csMasks[csid].setAllAssoc(parent.mm.addAss())
cmsk.merge(csMasks[csid])
parent.setMask(cmsk, setass=False)
def process(self):
parent = self.parent
weakTables = [
'formula', 'weight', 'totalcharge', 'xlogp', 'hbonddonor',
'hbondacceptor', 'heavyatom', 'tpsa'
]
list([parent.mm.setTableWeak(x) for x in weakTables])
ll = True
while ll:
ll = parent.getLine()
if not ll:
break
tmp = mask({})
tmp.append('_id', ll)
tmp2 = parent.mm.getMask(tmp, weak=True)
if not tmp2:
continue
else:
un = tmp2[0]
try:
pcMasks = self.pc.getPubchemMasks(un)
except:
print("#COMMENT unknown error")
continue
if parent.mm.debug:
print("#COMMENT mask " + str(ll))
msk = mask({})
for pcmsk in pcMasks:
pcmsk.setAllAssoc(parent.mm.addAss())
msk.merge(pcmsk)
parent.setMask(msk, setass=False)
def process(self):
""" parse the compunds.dat file from the Cyc databases and insert the
cycdb frame-id (uniqueid), synonyms, kegg and cas identifiers
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns) # new mask
un.MIN_OVERLAP = 1
ll = parent.getLine(comment='#')
while ll:
ll = ll.strip()
if ll == '//':
if un.nid() > 2:
try:
parent.setMask(un)
except:
# parent.mm.setMask(un, debug=True)
raise
un = mask({}, parent.mm.idpatterns)
un.MIN_OVERLAP = 1
# frameid
x = re.findall("UNIQUE-ID - (.*)", ll)
if x:
un.append(self.uniqueid, x[0], parent.confid, parent.sourceid)
# synonym
x = re.findall("COMMON-NAME - (.*)", ll)
if x:
un.append('synonym', x[0], parent.mm.confidence['weak'],
parent.sourceid)
x = re.findall("SYNONYMS - (.*)", ll)
if x:
un.append('synonym', x[0], parent.mm.confidence['weak'],
parent.sourceid)
# cas
x = re.findall("DBLINKS - \(CAS \"(\d{1,7}-\d{2}-\d{1})\"", ll)
if x:
un.append('cas', x[0], parent.confid, parent.sourceid)
# inchi
x = re.findall("INCHI - (.*)", ll)
if x:
un.append('inchi', x[0], parent.confid, parent.sourceid)
# smiles
x = re.findall("SMILES - (.*)", ll)
if x:
un.append('smiles', x[0], parent.mm.confidence['weak'],
parent.sourceid)
# kegg
x = re.findall("DBLINKS - \(LIGAND-CPD \"([A-Z][0-9]{5})\"", ll)
if x:
un.append('kegg', x[0], parent.confid, parent.sourceid)
ll = parent.getLine(comment='#')
if un.nid() > 2:
parent.setMask(un)
def process(self):
""" parse the Golm Metabolome Database standards library.
such as
those found at
http://csbdb.mpimp-golm.mpg.de/csbdb/gmd/msri/gmd_msri.html
"""
parent = self.parent
parent.mm.setTableWeak('mpimpclass')
# make sure that the necessary tables are in the db
un = mask({}, parent.mm.idpatterns)
l = parent.getLine()
while l:
# check and send off gathered data
if re.match('^Name:', l):
if un.nid(parent.mm) > 1:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
x = re.findall('Name: (.+)', l)
mpimpid = x[0].strip()
un.append('mpimp', mpimpid, parent.confid, parent.sourceid)
# get synonyms from the title, the other synonyms are
# just to full of crap
spam = mpimpid.split('_')
if spam[5].strip() != '':
un.append('synonym', spam[5], parent.confid,
parent.sourceid)
if re.match(".*\([0-9]+[tT][mM][sS]\)$", spam[5]):
syn = re.findall("(.*)\([0-9]+[tT][mM][sS]\)$",
spam[5])[0].strip()
un.append('synonym', syn, parent.confid,
parent.sourceid)
# ***** gather info *****
# mpimp class
x = re.findall('Synonym: Metabolite \(class\):(.+)', l)
if x:
un.append('mpimpclass', x[0].strip(),
parent.mm.confidence['weak'], parent.sourceid)
# cas 1
x = re.findall('CASNO: (.+)', l)
if x:
un.append('cas', x[0].strip(), parent.confid, parent.sourceid)
# cas 2
x = re.findall('CAS\|(\d{1,7}-\d{2}-\d{1})_\(.+\)', l)
if x:
un.append('cas', x[0].strip(), parent.confid, parent.sourceid)
# kegg
x = re.findall('KEGG\|([A-Z][0-9]{5})_.*\(.+\)', l)
if x:
un.append('kegg', x[0].strip(), parent.confid, parent.sourceid)
l = parent.getLine()
# send of the last mask if we have one
if un.nid(parent.mm) > 1:
parent.setMask(un)
def GetCompoundInfo(self, mm, csid):
"""
take an csid, get a mask
parameters:
-`mm`: a metmask database
-`csid`: a chemspider identifier
"""
tmpmask = mask({}, mm.idpatterns)
csid = self.parent.urlSafe(csid)
url = self.GetCompoundInfoURL + "CSID=" + str(csid) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return (tmpmask)
tmpmask.append('chemspider', csid, self.parent.confid, self.parent.sourceid)
searchResults = xml.dom.minidom.parse(qRes)
ids = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('CompoundInfo')))
inchi = list(mmquery.nodecontents(searchResults.getElementsByTagName('InChI')))
if inchi:
for ide in inchi:
tmpmask.append('inchi', ide, \
self.parent.confid, self.parent.sourceid)
smiles = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('SMILES')))
if smiles:
for ide in smiles:
tmpmask.append('smiles', ide, self.parent.confid, self.parent.sourceid)
inchikey = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('InChIKey')))
if inchikey:
for ide in inchikey:
tmpmask.append('inchikey', ide, self.parent.confid, self.parent.sourceid)
return (tmpmask)
def getCsMasks(self, un, mm):
""" take a mask and get filled mask from cs """
res = {}
tmpmask = mask({}, mm.idpatterns)
if not any([x in self.queryTables for x in un.getTables()]):
return (res)
if un.hasTable('chemspider'):
for csid in un.getIdentifiers('chemspider'):
res[csid] = self.GetCompoundInfo(mm, csid)
tmpmask.merge(res[csid])
def filteredids(table, un, tmpmask):
identifiers = []
if un.hasTable(table):
identifiers = un.getIdentifiers(table)
if tmpmask.hasTable(table):
identifiers = [x for x in identifiers if x \
not in tmpmask.getIdentifiers(table)]
return (identifiers)
def fillresandmerge(ide, tmpmask, res):
newids = self.SimpleSearch(ide)
for csid in newids:
if csid not in res:
res[csid] = self.GetCompoundInfo(mm, csid)
tmpmask.merge(res[csid])
for ide in filteredids('smiles', un, tmpmask):
fillresandmerge(ide, tmpmask, res)
return (res)
def getCsMasks(self, un, mm):
""" take a mask and get filled mask from cs """
res = {}
tmpmask = mask({}, mm.idpatterns)
if not any([x in self.queryTables for x in un.getTables()]):
return (res)
if un.hasTable('chemspider'):
for csid in un.getIdentifiers('chemspider'):
res[csid] = self.GetCompoundInfo(mm, csid)
tmpmask.merge(res[csid])
def filteredids(table, un, tmpmask):
identifiers = []
if un.hasTable(table):
identifiers = un.getIdentifiers(table)
if tmpmask.hasTable(table):
identifiers = [x for x in identifiers if x \
not in tmpmask.getIdentifiers(table)]
return (identifiers)
def fillresandmerge(ide, tmpmask, res):
newids = self.SimpleSearch(ide)
for csid in newids:
if csid not in res:
res[csid] = self.GetCompoundInfo(mm, csid)
tmpmask.merge(res[csid])
for ide in filteredids('smiles', un, tmpmask):
fillresandmerge(ide, tmpmask, res)
return (res)
def process (self) :
""" parse an SDF file and import any NISTID, synonyms, CASNO
and formula
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine()
parent.mm.setTableWeak('formula')
while ll :
if ll.startswith('$$$$') :
if un.nid(parent.mm) > 1 :
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
# synonym
if ll.startswith('> <NAME>') or ll.startswith('> <SYNONYMS>'):
tmp = parent.getLine().strip()
while tmp != "" :
un.append('synonym', tmp,
parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# nist
if ll.startswith('> <NISTNO>') :
tmp = parent.getLine().strip()
while tmp != "" :
un.append('nist', tmp,
parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# cas
if ll.startswith('> <CASNO>') :
tmp = parent.getLine().strip()
while tmp != "" :
un.append('cas', tmp,
parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# formula
if ll.startswith('> <FORMULA>') :
tmp = parent.getLine().strip()
while tmp != "" :
un.append('formula', tmp,
parent.mm.confidence['weak'], parent.sourceid)
tmp = parent.getLine().strip()
ll = parent.getLine()
if un.nid(parent.mm) > 1 :
parent.setMask(un)
def process(self):
""" parse an SDF file and import any NISTID, synonyms, CASNO
and formula
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine()
parent.mm.setTableWeak('formula')
while ll:
if ll.startswith('$$$$'):
if un.nid(parent.mm) > 1:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
# synonym
if ll.startswith('> <NAME>') or ll.startswith('> <SYNONYMS>'):
tmp = parent.getLine().strip()
while tmp != "":
un.append('synonym', tmp, parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# nist
if ll.startswith('> <NISTNO>'):
tmp = parent.getLine().strip()
while tmp != "":
un.append('nist', tmp, parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# cas
if ll.startswith('> <CASNO>'):
tmp = parent.getLine().strip()
while tmp != "":
un.append('cas', tmp, parent.confid, parent.sourceid)
tmp = parent.getLine().strip()
# formula
if ll.startswith('> <FORMULA>'):
tmp = parent.getLine().strip()
while tmp != "":
un.append('formula', tmp, parent.mm.confidence['weak'],
parent.sourceid)
tmp = parent.getLine().strip()
ll = parent.getLine()
if un.nid(parent.mm) > 1:
parent.setMask(un)
def process(self):
""" parse the RIKEN standards library file
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine()
parent.mm.setTableWeak('formula')
while ll:
x = re.findall('^Name: (.+)', ll)
if x:
if un.nid() > 1:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
un.append('riken', x[0].strip(),
parent.confid, parent.sourceid)
x = re.findall('CASNO: (\d{1,7}-\d{2}-\d{1})', ll)
if x:
un.append('cas', x[0].strip(),
parent.confid, parent.sourceid)
x = re.findall('^KEGG: ([A-Z][0-9]{5})', ll)
if x:
un.append('kegg', x[0].strip(),
parent.confid, parent.sourceid)
x = re.findall('^Formula: (.+)', ll)
if x:
un.append('formula', x[0].strip(),
parent.mm.confidence['weak'], parent.sourceid)
x = re.findall('^SMILES: (.+)', ll)
if x:
un.append('smiles', x[0].strip(),
parent.mm.confidence['weak'], parent.sourceid)
x = re.findall('Synonym: Name:(.+)', ll)
if x:
un.append('synonym', x[0].strip(),
parent.confid, parent.sourceid)
ll = parent.getLine()
if un.nid() > 1:
parent.setMask(un)
def process(self):
""" parse a file with metabocards from hmdb
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine(comment='\n')
if not ll.startswith("#BEGIN_METABOCARD"):
raise parserError("file does seem to contain metabocards: " + ll)
parent.mm.setTableWeak('formula')
ok = False
while ll:
if ll.startswith('#END_METABOCARD'):
if un.nid(parent.mm) > 1:
parent.setMask(un)
elif ll.startswith("#BEGIN_METABOCARD"):
un = mask({}, parent.mm.idpatterns)
elif ll.startswith("#"):
ok = False
if ll.strip() in self.tableDict:
con = self.tableDict[ll.strip()][0]
tab = self.tableDict[ll.strip()][1]
ok = True
else:
if ok:
identifiers = ll.split("; ")
for iden in identifiers:
if not iden.lower().startswith("not available"):
try:
un.append(tab, iden, con, parent.sourceid)
except idMisMatchError:
print("#OFFENDING LINE " + \
str(parent.lineNum) + "@" + \
tab + " : " + str(iden))
ll = parent.getLine(comment='\n')
def process(self):
""" parse a file with metabocards from hmdb
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine(comment='\n')
if not ll.startswith("#BEGIN_METABOCARD"):
raise parserError("file does seem to contain metabocards: " + ll)
parent.mm.setTableWeak('formula')
ok = False
while ll:
if ll.startswith('#END_METABOCARD'):
if un.nid(parent.mm) > 1:
parent.setMask(un)
elif ll.startswith("#BEGIN_METABOCARD"):
un = mask({}, parent.mm.idpatterns)
elif ll.startswith("#"):
ok = False
if ll.strip() in self.tableDict:
con = self.tableDict[ll.strip()][0]
tab = self.tableDict[ll.strip()][1]
ok = True
else:
if ok:
identifiers = ll.split("; ")
for iden in identifiers:
if not iden.lower().startswith("not available"):
try:
un.append(tab, iden, con, parent.sourceid)
except idMisMatchError:
print("#OFFENDING LINE " + \
str(parent.lineNum) + "@" + \
tab + " : " + str(iden))
ll = parent.getLine(comment='\n')
def test_insert_delete_mask(memory_db):
memory_db.createIdTable('table1')
memory_db.createIdTable('table2')
un = mask()
un.append('table1', 'foo', 'strong', 'test', 'baz')
un.append('table2', 'bar', 'strong', 'test', 'baz')
memory_db.setMask(un)
result = memory_db.simpleQuery('foo', 'table1', 'table2')
assert 'bar' in result[0][0]
memory_db.dropMask(un)
result = memory_db.simpleQuery('foo', 'table1', 'table2')
assert len(result[0][0]) == 0
memory_db.close()
assert not os.path.exists(':memory:')
def process(self):
""" parse the RIKEN standards library file
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns)
ll = parent.getLine()
parent.mm.setTableWeak('formula')
while ll:
x = re.findall('^Name: (.+)', ll)
if x:
if un.nid() > 1:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns)
un.append('riken', x[0].strip(), parent.confid,
parent.sourceid)
x = re.findall('CASNO: (\d{1,7}-\d{2}-\d{1})', ll)
if x:
un.append('cas', x[0].strip(), parent.confid, parent.sourceid)
x = re.findall('^KEGG: ([A-Z][0-9]{5})', ll)
if x:
un.append('kegg', x[0].strip(), parent.confid, parent.sourceid)
x = re.findall('^Formula: (.+)', ll)
if x:
un.append('formula', x[0].strip(),
parent.mm.confidence['weak'], parent.sourceid)
x = re.findall('^SMILES: (.+)', ll)
if x:
un.append('smiles', x[0].strip(), parent.mm.confidence['weak'],
parent.sourceid)
x = re.findall('Synonym: Name:(.+)', ll)
if x:
un.append('synonym', x[0].strip(), parent.confid,
parent.sourceid)
ll = parent.getLine()
if un.nid() > 1:
parent.setMask(un)
def process(self):
parent = self.parent
ncol = len(parent.tables)
## single identifier inserts should be allowed since they can
## might map to bins
# if ncol < 2:
# raise fileFormatError, "Only one column, pointless insertion"
# make sure that the necessary tables are in the db
ll = parent.getLine()
ll = ll.replace("#", "\\#")
while ll:
ll = ll.replace("#", "\\#")
try:
vec = fixLine(ll, sep=parent.sep1)
if len(vec) != ncol:
raise fileFormatError("Number of columns doesn't match the header :" \
+ str(parent.lineNum) + ll)
un = mask({}, parent.mm.idpatterns)
for i in range(0, len(vec)):
idvec = vec[i].strip().split(parent.sep2)
idvec[0].strip()
if idvec[0] and not re.match(parent.na, idvec[0]):
for ide in idvec:
ide = ide.strip()
if not re.match(parent.na, ide):
try:
un.append(parent.tables[i], ide,
self.confids[i],
parent.sourceid)
except idMisMatchError:
print("#OFFENDING LINE " + \
str(parent.lineNum) + "@" + \
parent.tables[i] + \
" : " + str(ide))
# no empty masks
if not un.isEmpty():
parent.setMask(un)
except KeyboardInterrupt:
raise Exception("Interrupt caught, breaking")
except fileFormatError:
print("#ERROR: format problem" + ll)
except:
print("#ERROR:" + ll)
ll = parent.getLine()
def process(self):
parent = self.parent
ncol = len(parent.tables)
## single identifier inserts should be allowed since they can
## might map to bins
# if ncol < 2:
# raise fileFormatError, "Only one column, pointless insertion"
# make sure that the necessary tables are in the db
ll = parent.getLine()
ll = ll.replace("#", "\\#")
while ll:
ll = ll.replace("#", "\\#")
try:
vec = fixLine(ll, sep=parent.sep1)
if len(vec) != ncol:
raise fileFormatError("Number of columns doesn't match the header :" \
+ str(parent.lineNum) + ll)
un = mask({}, parent.mm.idpatterns)
for i in range(0, len(vec)):
idvec = vec[i].strip().split(parent.sep2)
idvec[0].strip()
if idvec[0] and not re.match(parent.na, idvec[0]):
for ide in idvec:
ide = ide.strip()
if not re.match(parent.na, ide):
try:
un.append(parent.tables[i], ide,
self.confids[i], parent.sourceid)
except idMisMatchError:
print("#OFFENDING LINE " + \
str(parent.lineNum) + "@" + \
parent.tables[i] + \
" : " + str(ide))
# no empty masks
if not un.isEmpty():
parent.setMask(un)
except KeyboardInterrupt:
raise Exception("Interrupt caught, breaking")
except fileFormatError:
print("#ERROR: format problem" + ll)
except:
print("#ERROR:" + ll)
ll = parent.getLine()
def GetCompoundInfo(self, mm, csid):
"""
take an csid, get a mask
parameters:
-`mm`: a metmask database
-`csid`: a chemspider identifier
"""
tmpmask = mask({}, mm.idpatterns)
csid = self.parent.urlSafe(csid)
url = self.GetCompoundInfoURL + "CSID=" + str(
csid) + "&token=" + self.token
qRes = self.parent.getUrl(url)
if not qRes:
return (tmpmask)
tmpmask.append('chemspider', csid, self.parent.confid,
self.parent.sourceid)
searchResults = xml.dom.minidom.parse(qRes)
ids = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('CompoundInfo')))
inchi = list(
mmquery.nodecontents(searchResults.getElementsByTagName('InChI')))
if inchi:
for ide in inchi:
tmpmask.append('inchi', ide, \
self.parent.confid, self.parent.sourceid)
smiles = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('SMILES')))
if smiles:
for ide in smiles:
tmpmask.append('smiles', ide, self.parent.confid,
self.parent.sourceid)
inchikey = \
list(mmquery.nodecontents(searchResults.getElementsByTagName('InChIKey')))
if inchikey:
for ide in inchikey:
tmpmask.append('inchikey', ide, self.parent.confid,
self.parent.sourceid)
return (tmpmask)
def process(self):
""" parse a cyc dump file and fectch cas, kegg, synonym and
cycpath
"""
parent = self.parent
last = ""
parent.mm.setTableWeak('cycpath')
parent.mm.setTableWeak('formula')
l = parent.getLine()
while l:
l = l.replace("'", "\\'") # cyc likes apostrophes
l = l.replace("#", "\#") # hashes in the middle of lines
vec = fixLine(l, sep=parent.sep1)
if len(vec) < 6:
l = parent.getLine()
continue
if len(vec) == 9:
vec = [vec[0]] + [vec[1]] + [vec[5]] + [vec[8].strip()] + [vec[3]] + [vec[4]]
else:
vec = [vec[0]] + [vec[1]] + [vec[5]] + [''] + [vec[3]] + [vec[4]]
if vec == last:
l = parent.getLine()
continue
un = mask({}, parent.mm.idpatterns)
if len(vec) == 6:
# pointless if we dont have links
if vec[2] != '':
# cas
x = re.findall('CAS:(\d{1,7}-\d{2}-\d{1})',
vec[2])
for casno in x:
un.append('cas', casno, parent.confid,
parent.sourceid)
# kegg
x = re.findall('LIGAND-CPD:([A-Z][0-9]{5})',
vec[2])
for kegg in x:
un.append('kegg', kegg, parent.confid,
parent.sourceid)
# pubchem
x = re.findall('PUBCHEM:([0-9]+)',
vec[2])
for kegg in x:
un.append('pubchem', kegg, parent.confid,
parent.sourceid)
# knapsack
x = re.findall('KNAPSACK:([0-9]+)',
vec[2])
for kegg in x:
un.append('knapsack', kegg, parent.confid,
parent.sourceid)
# formula
if vec[4] != '':
un.append('formula', vec[4].replace(" ", ""),
parent.mm.confidence['weak'],
parent.sourceid)
# smiles
if vec[5] != '':
un.append('smiles', vec[5],
parent.mm.confidence['weak'],
parent.sourceid)
# synonym
un.append('synonym', vec[0], parent.mm.confidence['weak'],
parent.sourceid)
if vec[1] != '':
x = vec[1].split("*")
for syn in x:
un.append('synonym', syn,
parent.mm.confidence['weak'],
parent.sourceid)
if vec[3] != '':
un.append('cycpath', vec[3],
parent.mm.confidence['weak'],
parent.sourceid)
if un.nid() < 2 or \
all([not un.hasTable('cas'), not un.hasTable('kegg')]):
un = mask({}, parent.mm.idpatterns)
if un.nid() > 1:
parent.setMask(un)
last = vec
l = parent.getLine()
def pubchem2mask(self, docSum):
""" turn a docsum node in to a mask
"""
un = mask({}, self.parent.mm.idpatterns)
cid = next(mmquery.nodecontents(docSum.getElementsByTagName("Id")))
un.append('cid', cid, self.parent.confid, self.parent.sourceid)
p = re.compile('(<a href[^>]*>)|(</a>)|(ligand)|(,)')
cnf = self.parent.confid
src = self.parent.sourceid
weak = self.parent.mm.confidence['weak']
for item in docSum.getElementsByTagName('Item'):
if len(item.childNodes) == 0:
continue
nVal = item.childNodes[0].nodeValue
if item.getAttribute('Name') == 'SynonymList':
synonyms = list(mmquery.nodecontents(item.childNodes))
for s in synonyms:
s = p.sub('', s.strip())
if 'kegg' in guessTable(s, self.parent.mm.idpatterns):
un.append('kegg', s, cnf, src)
continue
if 'cas' in guessTable(s, self.parent.mm.idpatterns):
un.append('cas', s, cnf, src)
continue
if 'chebi' in guessTable(s, self.parent.mm.idpatterns):
un.append('chebi', s, cnf, src)
continue
if 'inchi' in guessTable(s, self.parent.mm.idpatterns):
un.append('inchi', s, cnf, src)
continue
else:
un.append('synonym', s, weak, src)
elif item.getAttribute('Name') == 'IUPACName':
un.append('iupac', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'InChIKey':
un.append('inchikey', nVal, cnf, src)
# annotation section
elif item.getAttribute('Name') == 'MolecularFormula':
un.append('formula', nVal, weak, src)
elif item.getAttribute('Name') == 'MolecularWeight':
un.append('weight', nVal, weak, src)
elif item.getAttribute('Name') == 'TotalFormalCharge':
un.append('totalcharge', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondDonorCount':
un.append('hbonddonor', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondAcceptorCount':
un.append('hbondacceptor', nVal, weak, src)
elif item.getAttribute('Name') == 'HeavyAtomCount':
un.append('heavyatom', nVal, weak, src)
elif item.getAttribute('Name') == 'TPSA':
un.append('tpsa', nVal, weak, src)
return (un)
def chebi2mask(self, mm, chebiId):
"""
get a mask containing the info associated with a chebi id
2. use getComplete to fetch the contents of the relevant entries
"""
ba = chebiId
un = mask({}, mm.idpatterns)
qUrl = self.getComplete + "chebiId=" + str(chebiId)
qRes = self.parent.getUrl(qUrl)
if not qRes:
return (mask({}))
searchResults = xml.dom.minidom.parse(qRes)
if not searchResults:
return (mask({}))
if searchResults.getElementsByTagName('ns1:return'):
retList = searchResults.getElementsByTagName('ns1:return')[0]
# found non-existent chebiId
else:
# this would have been reasonable but chebi has problems so that some
# entries exist, but cant be fetched. valid chebi but we cant query
# for it,
# hence, skip completely
# delmask = mask({})
# delmask.append('chebi', chebiId)
# mm.brandish(delmask)
return (mask({}))
# chebiid
if retList.getElementsByTagName("ns1:chebiId"):
chebiId = list(query.nodecontents(retList.getElementsByTagName("ns1:chebiId")))[0]
un.append('chebi', chebiId.replace("CHEBI:", ""),
self.parent.confid, self.parent.sourceid)
# smiles
if retList.getElementsByTagName("ns1:smiles"):
un.append('smiles', list(query.nodecontents(retList.getElementsByTagName("ns1:smiles")))[0],
self.parent.confid, self.parent.sourceid)
# synonym
if searchResults.getElementsByTagName('ns1:Synonyms'):
syns = searchResults.getElementsByTagName('ns1:Synonyms')[0]
for sy in list(query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('synonym', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
# inchi
if retList.getElementsByTagName("ns1:inchi"):
un.append('inchi', list(query.nodecontents(retList.getElementsByTagName("ns1:inchi")))[0],
self.parent.confid, self.parent.sourceid)
# iupac
if searchResults.getElementsByTagName('ns1:IupacNames'):
syns = searchResults.getElementsByTagName('ns1:IupacNames')[0]
for sy in list(query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('iupac', sy, self.parent.confid, \
self.parent.sourceid)
# kegg
for ll in searchResults.getElementsByTagName('ns1:DatabaseLinks'):
if list(query.nodecontents(ll.getElementsByTagName("ns1:type")))[0] == 'KEGG COMPOUND accession':
for sy in list(query.nodecontents(ll.getElementsByTagName("ns1:data"))):
un.append('kegg', sy, self.parent.confid, self.parent.sourceid)
# cas
for ll in searchResults.getElementsByTagName('ns1:RegistryNumbers'):
if list(query.nodecontents(ll.getElementsByTagName("ns1:type")))[0] == 'CAS Registry Number':
for sy in list(query.nodecontents(ll.getElementsByTagName("ns1:data"))):
un.append('cas', sy, self.parent.confid, self.parent.sourceid)
# pdb.set_trace()
# formula
if searchResults.getElementsByTagName('ns1:Formulae'):
form = searchResults.getElementsByTagName('ns1:Formulae')[0]
for sy in list(query.nodecontents(form.getElementsByTagName("ns1:data"))):
un.append('formula', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
return (un)
def getChebiMasks(self, un, mm):
"""
1. use getLite to query for all relevant entries in the mask
2. Query for chebi, iupac, cas, kegg, inchi, smiles
1. Get only exact matches
3. Take the unique ones
4. Get all is_a children of the found chebis and add them
Result is a list of masks
"""
res = {}
# pdb.set_trace()
tmpmask = mask({}, mm.idpatterns)
if not any([x in self.queryTables for x in un.getTables()]):
return (res)
if un.hasTable('chebi'):
for ch in un.getIdentifiers('chebi'):
res[ch] = self.chebi2mask(mm, ch)
tmpmask.merge(res[ch])
def filteredids(table, un, tmpmask):
identifiers = []
if un.hasTable(table):
identifiers = un.getIdentifiers(table)
if tmpmask.hasTable(table):
# to start querying for the same identifiers several times
identifiers = [x for x in identifiers if x not in tmpmask.getIdentifiers(table)]
return (identifiers)
def fillresandmerge(qUrl, tmpmask, res):
newids = self.url2ids(qUrl)
for ch in newids:
if ch not in res:
res[ch] = self.chebi2mask(mm, ch)
tmpmask.merge(res[ch])
for ide in filteredids('iupac', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=IUPAC+NAME"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('cas', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=REGISTRY+NUMBER"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('kegg', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=DATABASE+LINK"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('inchi', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=INCHI"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('smiles', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=SMILES"
fillresandmerge(qUrl, tmpmask, res)
foundIds = list(res.keys())
children = []
for ch in foundIds:
children = self.getChebiChildren(ch)
for child in children:
if child not in res:
res[child] = self.chebi2mask(mm, child)
return (res)
def process(self):
""" parse the compunds.dat file from the Cyc databases and insert the
cycdb frame-id (uniqueid), synonyms, kegg and cas identifiers
"""
parent = self.parent
un = mask({}, parent.mm.idpatterns) # new mask
un.MIN_OVERLAP = 1
ll = parent.getLine(comment='#')
while ll:
ll = ll.strip()
if ll == '//':
if un.nid() > 2:
try:
parent.setMask(un)
except:
# parent.mm.setMask(un, debug=True)
raise
un = mask({}, parent.mm.idpatterns)
un.MIN_OVERLAP = 1
# frameid
x = re.findall("UNIQUE-ID - (.*)", ll)
if x:
un.append(self.uniqueid, x[0], parent.confid,
parent.sourceid)
# synonym
x = re.findall("COMMON-NAME - (.*)", ll)
if x:
un.append('synonym', x[0],
parent.mm.confidence['weak'],
parent.sourceid)
x = re.findall("SYNONYMS - (.*)", ll)
if x:
un.append('synonym', x[0],
parent.mm.confidence['weak'],
parent.sourceid)
# cas
x = re.findall("DBLINKS - \(CAS \"(\d{1,7}-\d{2}-\d{1})\"", ll)
if x:
un.append('cas', x[0], parent.confid,
parent.sourceid)
# inchi
x = re.findall("INCHI - (.*)", ll)
if x:
un.append('inchi', x[0], parent.confid,
parent.sourceid)
# smiles
x = re.findall("SMILES - (.*)", ll)
if x:
un.append('smiles', x[0], parent.mm.confidence['weak'],
parent.sourceid)
# kegg
x = re.findall("DBLINKS - \(LIGAND-CPD \"([A-Z][0-9]{5})\"", ll)
if x:
un.append('kegg', x[0], parent.confid,
parent.sourceid)
ll = parent.getLine(comment='#')
if un.nid() > 2:
parent.setMask(un)
def getChebiMasks(self, un, mm):
"""
1. use getLite to query for all relevant entries in the mask
2. Query for chebi, iupac, cas, kegg, inchi, smiles
1. Get only exact matches
3. Take the unique ones
4. Get all is_a children of the found chebis and add them
Result is a list of masks
"""
res = {}
# pdb.set_trace()
tmpmask = mask({}, mm.idpatterns)
if not any([x in self.queryTables for x in un.getTables()]):
return (res)
if un.hasTable('chebi'):
for ch in un.getIdentifiers('chebi'):
res[ch] = self.chebi2mask(mm, ch)
tmpmask.merge(res[ch])
def filteredids(table, un, tmpmask):
identifiers = []
if un.hasTable(table):
identifiers = un.getIdentifiers(table)
if tmpmask.hasTable(table):
# to start querying for the same identifiers several times
identifiers = [
x for x in identifiers
if x not in tmpmask.getIdentifiers(table)
]
return (identifiers)
def fillresandmerge(qUrl, tmpmask, res):
newids = self.url2ids(qUrl)
for ch in newids:
if ch not in res:
res[ch] = self.chebi2mask(mm, ch)
tmpmask.merge(res[ch])
for ide in filteredids('iupac', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=IUPAC+NAME"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('cas', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=REGISTRY+NUMBER"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('kegg', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=DATABASE+LINK"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('inchi', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=INCHI"
fillresandmerge(qUrl, tmpmask, res)
for ide in filteredids('smiles', un, tmpmask):
qUrl = self.getLite + "search=" + \
self.parent.urlSafe(ide) + "&searchCategory=SMILES"
fillresandmerge(qUrl, tmpmask, res)
foundIds = list(res.keys())
children = []
for ch in foundIds:
children = self.getChebiChildren(ch)
for child in children:
if child not in res:
res[child] = self.chebi2mask(mm, child)
return (res)
def process(self):
""" parse a cyc dump file and fectch cas, kegg, synonym and
cycpath
"""
parent = self.parent
last = ""
parent.mm.setTableWeak('cycpath')
parent.mm.setTableWeak('formula')
l = parent.getLine()
while l:
l = l.replace("'", "\\'") # cyc likes apostrophes
l = l.replace("#", "\#") # hashes in the middle of lines
vec = fixLine(l, sep=parent.sep1)
if len(vec) < 6:
l = parent.getLine()
continue
if len(vec) == 9:
vec = [vec[0]] + [vec[1]] + [vec[5]] + [vec[8].strip()] + [
vec[3]
] + [vec[4]]
else:
vec = [vec[0]] + [vec[1]] + [vec[5]] + [''] + [vec[3]
] + [vec[4]]
if vec == last:
l = parent.getLine()
continue
un = mask({}, parent.mm.idpatterns)
if len(vec) == 6:
# pointless if we dont have links
if vec[2] != '':
# cas
x = re.findall('CAS:(\d{1,7}-\d{2}-\d{1})', vec[2])
for casno in x:
un.append('cas', casno, parent.confid, parent.sourceid)
# kegg
x = re.findall('LIGAND-CPD:([A-Z][0-9]{5})', vec[2])
for kegg in x:
un.append('kegg', kegg, parent.confid, parent.sourceid)
# pubchem
x = re.findall('PUBCHEM:([0-9]+)', vec[2])
for kegg in x:
un.append('pubchem', kegg, parent.confid,
parent.sourceid)
# knapsack
x = re.findall('KNAPSACK:([0-9]+)', vec[2])
for kegg in x:
un.append('knapsack', kegg, parent.confid,
parent.sourceid)
# formula
if vec[4] != '':
un.append('formula', vec[4].replace(" ", ""),
parent.mm.confidence['weak'],
parent.sourceid)
# smiles
if vec[5] != '':
un.append('smiles', vec[5],
parent.mm.confidence['weak'],
parent.sourceid)
# synonym
un.append('synonym', vec[0], parent.mm.confidence['weak'],
parent.sourceid)
if vec[1] != '':
x = vec[1].split("*")
for syn in x:
un.append('synonym', syn,
parent.mm.confidence['weak'],
parent.sourceid)
if vec[3] != '':
un.append('cycpath', vec[3],
parent.mm.confidence['weak'],
parent.sourceid)
if un.nid() < 2 or \
all([not un.hasTable('cas'), not un.hasTable('kegg')]):
un = mask({}, parent.mm.idpatterns)
if un.nid() > 1:
parent.setMask(un)
last = vec
l = parent.getLine()
def chebi2mask(self, mm, chebiId):
"""
get a mask containing the info associated with a chebi id
2. use getComplete to fetch the contents of the relevant entries
"""
ba = chebiId
un = mask({}, mm.idpatterns)
qUrl = self.getComplete + "chebiId=" + str(chebiId)
qRes = self.parent.getUrl(qUrl)
if not qRes:
return (mask({}))
searchResults = xml.dom.minidom.parse(qRes)
if not searchResults:
return (mask({}))
if searchResults.getElementsByTagName('ns1:return'):
retList = searchResults.getElementsByTagName('ns1:return')[0]
# found non-existent chebiId
else:
# this would have been reasonable but chebi has problems so that some
# entries exist, but cant be fetched. valid chebi but we cant query
# for it,
# hence, skip completely
# delmask = mask({})
# delmask.append('chebi', chebiId)
# mm.brandish(delmask)
return (mask({}))
# chebiid
if retList.getElementsByTagName("ns1:chebiId"):
chebiId = list(
query.nodecontents(
retList.getElementsByTagName("ns1:chebiId")))[0]
un.append('chebi', chebiId.replace("CHEBI:", ""),
self.parent.confid, self.parent.sourceid)
# smiles
if retList.getElementsByTagName("ns1:smiles"):
un.append(
'smiles',
list(
query.nodecontents(
retList.getElementsByTagName("ns1:smiles")))[0],
self.parent.confid, self.parent.sourceid)
# synonym
if searchResults.getElementsByTagName('ns1:Synonyms'):
syns = searchResults.getElementsByTagName('ns1:Synonyms')[0]
for sy in list(
query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('synonym', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
# inchi
if retList.getElementsByTagName("ns1:inchi"):
un.append(
'inchi',
list(
query.nodecontents(
retList.getElementsByTagName("ns1:inchi")))[0],
self.parent.confid, self.parent.sourceid)
# iupac
if searchResults.getElementsByTagName('ns1:IupacNames'):
syns = searchResults.getElementsByTagName('ns1:IupacNames')[0]
for sy in list(
query.nodecontents(syns.getElementsByTagName("ns1:data"))):
un.append('iupac', sy, self.parent.confid, \
self.parent.sourceid)
# kegg
for ll in searchResults.getElementsByTagName('ns1:DatabaseLinks'):
if list(query.nodecontents(ll.getElementsByTagName(
"ns1:type")))[0] == 'KEGG COMPOUND accession':
for sy in list(
query.nodecontents(
ll.getElementsByTagName("ns1:data"))):
un.append('kegg', sy, self.parent.confid,
self.parent.sourceid)
# cas
for ll in searchResults.getElementsByTagName('ns1:RegistryNumbers'):
if list(query.nodecontents(ll.getElementsByTagName(
"ns1:type")))[0] == 'CAS Registry Number':
for sy in list(
query.nodecontents(
ll.getElementsByTagName("ns1:data"))):
un.append('cas', sy, self.parent.confid,
self.parent.sourceid)
# pdb.set_trace()
# formula
if searchResults.getElementsByTagName('ns1:Formulae'):
form = searchResults.getElementsByTagName('ns1:Formulae')[0]
for sy in list(
query.nodecontents(form.getElementsByTagName("ns1:data"))):
un.append('formula', sy, self.parent.mm.confidence['weak'], \
self.parent.sourceid)
return (un)
def process(self):
parent = self.parent
# annotate this table as fully weak
parent.mm.setTableWeak('pathway')
parent.mm.setTableWeak('formula')
ll = parent.getLine()
if not re.match("^ENTRY", ll):
raise fileFormatError("COMPOUNDS file doesn't look like expected")
un = mask({}, parent.mm.idpatterns)
good = True
while ll:
# check and send off gathered data
# kegg
x = re.findall('ENTRY *([A-Z][0-9]{5}) ', ll)
if x:
if not un.isEmpty() and good:
parent.setMask(un)
un = mask({}, parent.mm.idpatterns) # new mask
keggno = x[0].strip()
un.append('kegg', keggno, parent.confid,
parent.sourceid)
good = True
# synonym
x = re.findall('NAME +(.+)', ll)
if x:
un.append('synonym',
x[0].strip().replace(";", ""),
parent.confid, parent.sourceid)
while re.match(".+;$", ll):
ll = parent.getLine()
un.append('synonym',
ll.strip().replace(";", ""),
parent.confid,
parent.sourceid)
# formula
x = re.findall('FORMULA +(.+)', ll)
if x:
un.append('formula', x[0], parent.mm.confidence['weak'],
parent.sourceid)
# knapsack
x = re.findall('KNApSAcK: (.+)', ll)
if x:
un.append('knapsack', x[0], parent.confid,
parent.sourceid)
# cas
if re.findall('CAS:', ll):
x = re.findall('(\d{2,7}-\d{2}-\d{1})', ll)
for xx in x:
un.append('cas', xx, parent.confid,
parent.sourceid)
# sid <- obs not cid!
x = re.findall('PubChem: (\d+)', ll)
if x:
un.append('sid', x[0], parent.confid,
parent.sourceid)
# chebi
x = re.findall('ChEBI: (\d+)', ll)
if x:
un.append('chebi', x[0], parent.confid,
parent.sourceid)
# pathway
x = re.findall('(PATHWAY)*\s*(ko|map)(\d+)', ll)
if x:
un.append('pathway', x[0][2],
parent.mm.confidence['weak'],
parent.sourceid)
# comment, to get rid of kegg internal generics
if re.match('ENTRY *[A-Z][0-9]{5}$ *Peptide *Compound$', ll):
good = False
if re.match('COMMENT', ll):
if re.match('COMMENT generic compound in reaction hierarchy', ll) or \
re.match('COMMENT coordination compound', ll) or \
re.match('COMMENT.+R=.+', ll):
good = False
ll = parent.getLine()
# send of the last mask if we have one
if not un.isEmpty() and good:
parent.setMask(un)
def pubchem2mask(self, docSum):
""" turn a docsum node in to a mask
"""
un = mask({}, self.parent.mm.idpatterns)
cid = next(mmquery.nodecontents(docSum.getElementsByTagName("Id")))
un.append('cid', cid, self.parent.confid, self.parent.sourceid)
p = re.compile('(<a href[^>]*>)|(</a>)|(ligand)|(,)')
cnf = self.parent.confid
src = self.parent.sourceid
weak = self.parent.mm.confidence['weak']
for item in docSum.getElementsByTagName('Item'):
if len(item.childNodes) == 0:
continue
nVal = item.childNodes[0].nodeValue
if item.getAttribute('Name') == 'SynonymList':
synonyms = list(mmquery.nodecontents(item.childNodes))
for s in synonyms:
s = p.sub('', s.strip())
if 'kegg' in guessTable(s, self.parent.mm.idpatterns):
un.append('kegg', s, cnf, src)
continue
if 'cas' in guessTable(s, self.parent.mm.idpatterns):
un.append('cas', s, cnf, src)
continue
if 'chebi' in guessTable(s, self.parent.mm.idpatterns):
un.append('chebi', s, cnf, src)
continue
if 'inchi' in guessTable(s, self.parent.mm.idpatterns):
un.append('inchi', s, cnf, src)
continue
else:
un.append('synonym', s, weak, src)
elif item.getAttribute('Name') == 'IUPACName':
un.append('iupac', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'CanonicalSmile':
un.append('smiles', nVal, cnf, src)
elif item.getAttribute('Name') == 'InChIKey':
un.append('inchikey', nVal, cnf, src)
# annotation section
elif item.getAttribute('Name') == 'MolecularFormula':
un.append('formula', nVal, weak, src)
elif item.getAttribute('Name') == 'MolecularWeight':
un.append('weight', nVal, weak, src)
elif item.getAttribute('Name') == 'TotalFormalCharge':
un.append('totalcharge', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'XLogP':
un.append('xlogp', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondDonorCount':
un.append('hbonddonor', nVal, weak, src)
elif item.getAttribute('Name') == 'HydrogenBondAcceptorCount':
un.append('hbondacceptor', nVal, weak, src)
elif item.getAttribute('Name') == 'HeavyAtomCount':
un.append('heavyatom', nVal, weak, src)
elif item.getAttribute('Name') == 'TPSA':
un.append('tpsa', nVal, weak, src)
return (un)